Agricultural structured data acquisition method and system based on voice intelligence

Abstract

The application relates to the field of artificial intelligence, in particular to an agricultural structured data acquisition method and system based on voice intelligence, which determines an acquisition scene and calls a predefined field set by receiving voice input in the field of agriculture; the voice recognition text is intelligently matched with the field set by using the deep semantic understanding capability of a large language model, accurate mapping and standardized conversion of colloquial expressions, professional terms and numerical units are carried out, and initial structured data is generated; further, the conversion result driven by the large model is logically checked and reasonably analyzed in combination with preset agricultural field knowledge rules, data confirmation or correction is realized through voice interaction, and finally high-quality structured data is output. The cognitive reasoning capability of the large language model is deeply integrated into the whole link of agricultural data acquisition, intelligent conversion from non-professional voice to standardized data is realized, and the problems of low efficiency of traditional methods and lack of field semantic understanding of general voice recognition are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to an agricultural structured data acquisition method and system based on voice intelligence. BACKGROUND

[0002] In current agricultural research and production management, it is necessary to systematically collect multiple types of information such as crop traits, environmental parameters, and farming operations, and organize them into agricultural structured data, i.e., table-type data with clear field definitions, standard data types, and standard units, which can be directly used for statistical analysis and intelligent decision-making. Currently, the generation of such structured data mainly relies on two ways: one is for workers to record on paper forms in the field and then manually enter them into computers after returning, which is a cumbersome process prone to errors and low in timeliness; the other is to use electronic forms on mobile terminals such as smartphones for on-site point selection or input, which improves the level of digitization but makes screen operation difficult in typical agricultural environments such as strong light, rain, and muddy hands, and the actual input efficiency is still not ideal.

[0003] In the prior art, at the data entry and preliminary structuring level, the prior art cannot achieve the automatic conversion of "semantic understanding" and "field adaptation". Specifically, general voice recognition technology can only convert voice into text, but cannot understand the specific business scenarios and data connotations of agriculture. For example, it cannot associate "piglet" with the standard field "plant height", and the existing method is limited to the format or simple range of the field itself, lacking the cross-field logical verification based on agricultural domain knowledge, such as the ability to identify contradictions such as "sowing period plant height 3 meters" that violate common sense of farming.

[0004] Secondly, at the agricultural professional application level, the prior art lacks adaptability to the complexity and dynamics of agricultural production, resulting in insufficient data rationality and usability. The data of crops or livestock and poultry changes dynamically with growth stages, breed characteristics, and environmental conditions. The existing rigid data verification model may not be able to distinguish between "measurement errors" and "environmentally induced reasonable abnormalities", and may misjudge valuable stress physiology data as invalid data and filter them out. SUMMARY

[0005] In order to solve one or more problems in the prior art, the main purpose of the present application is to provide an agricultural structured data acquisition method and system based on voice intelligence.

[0006] In order to achieve the above-mentioned purpose of the application, the present application provides an agricultural structured data acquisition method based on voice intelligence, which comprises:

[0007] receiving user input of agricultural field-related voice data;

[0008] Based on the voice data, the current data acquisition scenario is determined, and a predefined set of acquisition fields for the acquisition scenario is invoked, wherein the set of acquisition fields includes the standard name, data type, and value constraints of the fields;

[0009] Perform text information recognition on the voice data;

[0010] Based on the recognition results, the recognized text information is matched with the collection field set, the expression in the text is mapped to the corresponding standard field name, and the numerical value and unit in the text are standardized and converted according to the data type and value constraints to generate initial structured data.

[0011] The initial structured data is validated based on the value constraints in the collection field set and the preset agricultural knowledge rules.

[0012] Based on the verification results, the final structured data is determined.

[0013] This application also provides a voice-based intelligent agricultural structured data acquisition system, including:

[0014] The receiving module is used to receive voice data related to agriculture input by the user;

[0015] The calling module is used to determine the current data acquisition scenario based on the voice data, and to call a set of acquisition fields predefined in the acquisition scenario, wherein the set of acquisition fields includes the standard name, data type and value constraints of the fields;

[0016] The recognition module is used to perform text information recognition on the voice data;

[0017] The generation module is used to match the identified text information with the collection field set based on the recognition results, map the expressions in the text to the corresponding standard field names, and standardize and convert the numerical values ​​and units in the text according to the data type and value constraints to generate initial structured data.

[0018] The verification module is used to verify the initial structured data based on the value constraints in the collection field set and the preset agricultural knowledge rules.

[0019] The determination module is used to determine the final structured data based on the results of the verification.

[0020] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods described above.

[0021] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0022] This application's embodiment of the voice-based intelligent agricultural structured data acquisition method and system deeply integrates agricultural knowledge into the entire voice processing process, fundamentally solving the disconnect between voice recognition and the needs of agricultural data structuring in existing technologies. Firstly, the method frees the user's hands and eyes through pure voice interaction, greatly improving operational adaptability and data entry efficiency in harsh field environments. More importantly, it uses a scenario-based predefined field set as a "blueprint" to guide the system in intelligently mapping from free speech to standard fields and standardizing numerical units, achieving automatic and accurate conversion from unstructured speech to structured data, overcoming the challenges of semantic understanding and format unification. Furthermore, it introduces a verification mechanism based on agricultural knowledge rules to conduct in-depth quality checks on the data, including logical relationships, thereby ensuring the agronomical rationality and direct usability of the output data. Overall, this solution not only achieves a leap in acquisition efficiency but also, through the empowerment of domain knowledge, ensures the high quality and high reliability of agricultural data from the source acquisition stage. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating an embodiment of a voice-based intelligent agricultural structured data acquisition method according to this application.

[0024] Figure 2 This is a flowchart illustrating an embodiment of a voice-based intelligent agricultural structured data acquisition method according to this application.

[0025] Figure 3 This is a schematic block diagram of a voice-intelligent agricultural structured data acquisition system according to an embodiment of this application;

[0026] Figure 4 This is a schematic block diagram of the structure of a computer device according to an embodiment of this application.

[0027] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0029] Reference Figure 1This application provides a voice-based intelligent agricultural structured data acquisition method, the method comprising:

[0030] S1. Receive voice data related to agriculture input by the user;

[0031] S2. Based on the voice data, determine the current data acquisition scenario and call the collection field set predefined in the acquisition scenario, wherein the collection field set includes the standard name, data type and value constraints of the field;

[0032] S3. Perform text information recognition on the voice data;

[0033] S4. Based on the recognition results, the recognized text information is matched with the collection field set, the expression in the text is mapped to the corresponding standard field name, and the numerical value and unit in the text are standardized and converted according to the data type and value constraints to generate initial structured data.

[0034] S5. Based on the value constraints in the collection field set and the preset agricultural knowledge rules, the initial structured data is validated;

[0035] S6. Based on the verification results, determine the final structured data.

[0036] As described in steps S1-S3 above, step 1 captures the user's spoken voice signal in the agricultural field using audio input devices such as microphones, serving as the raw input for data processing. This fundamentally changes the data collection method, transforming the interaction mode from touchscreen or pen-and-paper interaction that relies on "hand-eye coordination" to pure voice interaction. This allows users to conveniently input data even when their hands are occupied, in harsh environments, or while moving, greatly improving the on-site adaptability of data collection. Step 2 performs preliminary analysis of the voice content, identifying keywords related to specific agricultural activities to determine the current task's scenario. Subsequently, the system calls a pre-bound, structured set of collection fields for that scenario as a "blueprint" for subsequent processing. This achieves "scenario-based" and "standardized" data collection. The predefined field set ensures the uniformity and integrity of the data structure, avoiding field omissions or confusion caused by users' arbitrary speech. Step 3 uses speech recognition technology to convert continuous voice signals into corresponding text sequences. This process can be completed based on a general or agricultural speech corpus-optimized recognition engine. Converting audio signals, which are difficult to calculate directly, into text information that can be processed by natural language processing technology is a prerequisite for all subsequent intelligent understanding and conversion.

[0037] As described in steps S4-S6 above, step 4 is the core conversion step. The system maps words in the text to corresponding fields based on the standard names and related word lists in the field set. Simultaneously, it intelligently converts and standardizes numerical values ​​and units in the text according to the data type and value constraints of the fields. This directly addresses a deficiency in existing technologies. It solves the problem of general speech recognition "only translating, not understanding," achieving: semantic error correction and normalization: automatically mapping colloquial or erroneous words such as "pig lamb" and "seedling height" to the standard field "plant height." A leap from unstructured to structured data: extracting multiple field values ​​from a sentence and converting descriptions such as "two meters three" into the standard numerical value "230" and the unit "cm," generating directly processable initial structured data, bridging the critical gap between speech and database. In step 5, after initial structuring, the system does not directly accept the data but applies two layers of verification: first, verification based on the value constraints of the fields themselves; second, verification of cross-field logical relationships using preset agricultural knowledge rules. Domain knowledge-driven quality control is introduced. This method not only checks the data format but also identifies logical contradictions that violate agricultural common sense, such as "plant height 3 meters at sowing time," significantly improving data reliability and agronomic rationality. Step 6 confirms the data that passes verification and marks or triggers subsequent correction processes for data that fails verification, ultimately outputting a set of structured data that has undergone quality control. This ensures the credibility and usability of the output data. The final generated dataset is formatted correctly and logically consistent, and can be directly imported into databases or analysis software, achieving a data-ready state where data collection is immediately followed by analysis. In this embodiment, this solution creatively injects agricultural knowledge into the speech conversion process through a combination of "scenario-based field set invocation" and "intelligent mapping and standardization based on field definitions." This is not merely "recognizing text" but also "understanding intent and formatting," achieving automated and highly accurate conversion from free speech to standard field values. By introducing a "verification based on agricultural knowledge rules" step, data quality control is upgraded from simple value range checks to correlation review containing agronomic logic. This makes it possible for the system to distinguish between "erroneous data" and "special agricultural conditions."

[0038] As mentioned above, deeply integrating agricultural knowledge into the entire voice processing workflow fundamentally solves the disconnect between existing technologies and the needs of structuring agricultural data. Firstly, the method frees the user's hands and eyes through pure voice interaction, greatly improving operational adaptability and data entry efficiency in harsh field environments. More importantly, by using a scenario-based predefined set of fields as a "blueprint," the system is guided to intelligently map free speech to standard fields and standardize numerical units, achieving automatic and accurate conversion from unstructured speech to structured data, overcoming the challenges of semantic understanding and format uniformity. Building on this, a verification mechanism based on agricultural knowledge rules is further introduced to conduct in-depth quality checks on the data, including logical relationships, thereby ensuring the agronomic rationality and direct usability of the output data. Overall, this solution not only achieves a leap in data collection efficiency but also, through the empowerment of domain knowledge, ensures the high quality and high reliability of agricultural data from the source collection stage.

[0039] Reference Figure 2 In one embodiment, the step of validating the initial structured data based on the value constraints in the collection field set and preset agricultural knowledge rules includes:

[0040] S51. Obtain the preset agricultural knowledge rules, wherein the agricultural knowledge rules define the logical constraint relationships between different collection fields;

[0041] S52. Substitute the values ​​of the relevant fields in the initial structured data into the corresponding logical constraint relationships for calculation;

[0042] S53. Determine whether the initial structured data satisfies the logical constraint relationship based on the calculation results;

[0043] S54. In response to the judgment being satisfied, determine that the initial structured data has passed the verification;

[0044] S55. In response to the determination that the condition is not met, the initial structured data is marked or a correction is triggered.

[0045] As described above, Step 1 retrieves specific agricultural business rules from a pre-built knowledge base. These rules are not isolated field requirements, but rather define the correlation conditions that multiple field values ​​should satisfy in the form of logical expressions. This transforms abstract domain knowledge into concrete logical relationships that can be recognized and executed by a computer. For example, the experience that "the ear position should generally not exceed 70% of the plant height" is transformed into a calculable rule, providing a clear basis for automated and in-depth verification. Step 2 extracts the corresponding specific values ​​from the generated initial structured data based on the fields involved in the rule, substitutes them into the logical expression for calculation, and obtains a Boolean value or quantitative result. This achieves the connection between knowledge and data. Through calculation, actual observation data is placed in the agricultural knowledge system for quantitative verification, making implicit experience explicit. Step 3 analyzes the calculation results of Step 2 and, based on preset judgment criteria, concludes whether the data passes the current rule's verification. This completes the automated decision-making process from calculation to judgment. The system can automatically identify data that is valid within a single field but violates cross-field agricultural logic, such as identifying contradictory records where the sowing period has the plant height at maturity. Steps 4 and 5 perform a data splitting operation based on the judgment conclusion: if the conditions are met, the data validity is confirmed; if not, an anomaly marker is added to the data, or a subsequent interactive correction process is triggered. This forms a closed-loop verification and data governance mechanism. It not only identifies problems but also drives their resolution, ensuring that all data entering the final dataset has passed the knowledge rule review or that abnormal data has been clearly marked, thus guaranteeing the overall quality and credibility of the dataset. In this embodiment, existing technologies typically only perform independent field-level verification and cannot capture the agronomical logical contradictions inherent between data fields. This solution, by introducing the definition and calculation judgment mechanism of "logical constraint relationships," can discover deep-seated errors such as "ear position higher than plant height" or "specific traits appearing at impossible growth stages," errors that traditional value range verification cannot reach.

[0046] In one embodiment, the steps of matching the identified text information with the collection field set, mapping the expressions in the text to corresponding standard field names, and standardizing and converting the numerical values ​​and units in the text according to the data type and value constraints to generate initial structured data include:

[0047] Based on the standard names of each field in the collection of fields and the predefined word list, the words in the text information are matched with the corresponding fields;

[0048] Extract the numerical and unit descriptions corresponding to each field from the matched text information;

[0049] Based on preset conversion rules, the numerical description is converted into a standard numerical format that conforms to the data type;

[0050] Based on the units defined for each field in the collection field set, the unit descriptions are converted or transformed into the defined units;

[0051] Based on the standard numerical format and the converted or transformed units, the values ​​of each field in the initial structured data are generated.

[0052] As described above, Step 1 utilizes a pre-built word list containing standard field names and their common colloquial expressions, aliases, and even common misidentified words to scan and semantically compare the identified text, mapping the words in the text to the standard fields defined in the collected field set. This achieves domain semantic understanding and error correction. It enables the system to understand that expressions such as "pig lamb" and "seedling height" truly refer to the "plant height" field, overcoming the fundamental obstacle of general speech recognition in mapping professional domain terms and completing the crucial alignment from raw text to business fields. Step 2, after determining the field to which the text fragment belongs, uses natural language processing technology to identify and separate words or phrases describing numerical values ​​and units of measurement from the fragment or adjacent context. This achieves precise location and extraction of key information. It structurally extracts the effective data components mixed in natural language, preparing raw materials for subsequent standardization processing. Step 3 calls a numerical conversion rule library designed for agricultural scenarios to process the extracted numerical descriptions. This includes converting Chinese numerals to Arabic numerals, quantifying fuzzy quantifiers, and standardizing decimal and percentage formats. This step solves the problem of non-standardized numerical values ​​in agricultural colloquial language. Whether the user says "one thousand two hundred" or "one thousand two," the system outputs a standard "1200," ensuring the mathematical computability and consistency of the data. Step 4 uses the clearly defined standard units for each field in the collected field set as a benchmark to identify and extract the unit descriptions, performing conversions when necessary. For example, the user's "mu" (acre) is converted to "square meters" as defined in the field according to conversion rules. This achieves uniformity and comparability of measurement units, eliminating unit confusion caused by different user habits and ensuring all data is based on the same measurement standard. Step 5 combines the fully standardized numerical values ​​with units as the final value for the field and fills it into the corresponding position in the initial structured data object. This completes the final assembly from unstructured text to machine-readable structured data. At this point, a data record with a standardized format, clear meaning, and direct usability for database operations or business logic is generated.

[0053] In one embodiment, after the step of determining the current data acquisition scenario, the method further includes:

[0054] Based on the growth stage or growth cycle of the target crop or livestock corresponding to the collection scenario, establish a physiological indicator mapping relationship that matches the growth stage or growth cycle.

[0055] Based on the physiological indicator mapping relationship, a rationality analysis is performed on the fields involving physiological or growth status in the initial structured data;

[0056] Based on the results of the rationality analysis, if the analysis results are unreasonable, candidate correction data are generated based on the physiological indicator mapping relationship, and the user is guided to confirm or correct the data through voice interaction.

[0057] As described above, after determining the data collection scenario, Step 1 further determines the specific growth stage of the organism based on the current time, variety information, or user specifications. Subsequently, it invokes the knowledge model corresponding to this stage, which defines the reasonable numerical range or typical characteristics of various physiological state fields at this stage. This is essentially a dynamic standard template, achieving a fundamental shift from static to dynamic verification standards. It enables the system to understand that the reasonable values ​​for "plant height" are completely different at the seedling, jointing, and maturity stages, thus providing a biologically sound and up-to-date judgment benchmark for subsequent analysis. Step 2 compares and evaluates the values ​​of fields describing growth traits in the initial structured data reported by the user with the dynamic standards established in Step 1 for the current growth stage. This analysis aims to determine whether the observed values ​​significantly deviate from the normal trajectory of its life cycle. A longitudinal rationality review based on the life cycle is introduced. This allows the system to discover phenomena such as "precocious" or "delayed development" that may appear normal in isolation but are abnormal when placed on a growth timeline, achieving deeper data insights. Step 3: When data is deemed unreasonable, the system does not simply reject it. Instead, it infers a more likely reasonable value as a candidate suggestion based on the mapping relationship of physiological indicators. This suggestion is then presented to the user via voice dialogue, guiding them to conduct on-site verification and decision-making. This upgrades data quality control from one-way alarm to two-way collaborative intelligent assistance. It not only identifies problems but also provides solutions using knowledge, reducing the user's operational burden through natural voice interaction, achieving efficient human-machine collaborative error correction in complex field environments. In this embodiment, the quality control dimension of agricultural data collection is expanded from static, field-to-field relationship verification to dynamic, time-series growth and development pattern verification. This addresses the shortcomings pointed out in the background technology: existing technologies lack adaptability to the complexity and dynamism of agricultural production, potentially leading to misjudgments of reasonable anomalies. Crop trait expression is the orderly unfolding of its inherent genetic program over time. By establishing a "growth stage-physiological indicator" mapping relationship, the system gains the ability to simulate this cognitive process for the first time. This allows verification to no longer focus solely on horizontal contradictions between data points but also examine whether individual data points are "misaligned" vertically along the growth curve. An intelligent verification layer with temporal awareness was constructed. This improved the system's ability to judge the rationality of data and provided a core framework for handling more complex scenarios (such as identifying the overall shift of the growth curve under environmental stress).

[0058] In one embodiment, the step of performing a rationality analysis on fields related to physiological or growth states in the initial structured data based on the physiological indicator mapping relationship includes:

[0059] From the physiological indicator mapping relationship, extract the reference value range or reference value feature of each physiological state field corresponding to the reproductive stage or growth cycle;

[0060] The actual values ​​of the same physiological state fields in the initial structured data are compared with the corresponding reference value range or reference value features.

[0061] Calculate the deviation between the actual value and the corresponding reference standard based on the comparison results;

[0062] Based on a preset deviation threshold, the reasonableness level of the actual value is determined, and the reasonableness level includes at least reasonable, questionable, and unreasonable.

[0063] As described above, Step 1, based on the currently determined reproductive stage, retrieves and obtains the quantitative reference standard for each physiological state field under this stage from the dynamic physiological indicator mapping relationship. This standard is not a fixed value, but a reasonable numerical range or a typical characteristic description. It transforms abstract agricultural growth laws into concrete and operable data benchmarks. For example, for the field of "plant height of wheat at the jointing stage," a specific range of "30 to 80 cm" is extracted, providing a precise and timely objective basis for subsequent quantitative comparison. Step 2 compares the actual observation values ​​reported by the user with the corresponding reference standard extracted in Step 1 mathematically or logically. This is a process of locating specific data points within a reasonable background range, achieving the first intersection of data and knowledge. Through comparison, it is possible to intuitively determine whether an observation value falls within a regular range, or is located on its edge or outside, laying the foundation for subsequent refined evaluation. Step 3 uses a preset algorithm to quantify the degree to which the actual value deviates from the reference standard. This can be calculating the distance from the range boundary or the relative error with the typical value, ultimately outputting a value representing the magnitude of the deviation. A precise quantitative evaluation mechanism is introduced. It transforms the vague perception of "too high" or "too low" into a precise measurement of "deviation of 20%", shifting the judgment of reasonableness from qualitative to quantitative and supporting more refined decision-making. Step 4: The system presets multiple deviation thresholds, forming a hierarchical judgment standard. Based on which threshold range the calculated deviation falls into, the reasonableness of the data is divided into multiple levels such as "reasonable," "questionable," or "unreasonable." This achieves hierarchical and flexible quality assessment. It changes the binary judgment mode of right or wrong, allowing the system to identify and distinguish different situations such as "completely normal," "slightly questionable but possibly reasonable," and "highly likely wrong," providing key decision-making basis for differentiated subsequent processing.

[0064] In one embodiment, the step of extracting reference value ranges or reference value features of each physiological state field corresponding to the reproductive stage or growth cycle from the physiological indicator mapping relationship includes:

[0065] Obtain current environmental status information associated with the target crop or target livestock;

[0066] Based on the current environmental state information, it is determined whether there is a predefined stress environment type. The stress environment type is used to characterize the set of environmental conditions that have a specific pattern of influence on the growth and development of the target crop or target livestock.

[0067] If it exists, then extract the reference value range or reference value feature that corresponds to the reproductive stage or growth cycle and the stress environment type from the physiological indicator mapping relationship;

[0068] If not, then extract the reference value range or reference value feature corresponding to the reproductive stage or growth cycle under normal conditions from the physiological indicator mapping relationship.

[0069] As described above, step 1 acquires environmental data from the collection site in real-time or near real-time by connecting to IoT devices such as weather stations and soil sensors, or through brief verbal descriptions by the user. This data constitutes the key context for assessing the growth background. It endows the system with environmental perception capabilities, enabling its judgments to extend beyond the organism's own data and place observations within a specific environmental context, providing crucial input for contextualized analysis. Step 2 matches the acquired environmental data with various stress condition models defined in the knowledge base. These models describe a set of patterned environmental conditions that have specific adverse effects on growth and development, such as continuous drought, persistent low temperatures, or nutrient deficiencies. This achieves automated identification and classification of adverse conditions. Like an agronomical expert, the system can identify types of stress that may pose substantial threats to crops from a wealth of environmental data, thereby initiating corresponding intelligent processing logic. Steps 3 and 4 constitute a key branch logic. If a specific stress is identified, the system retrieves a reference standard from the knowledge base for that growth stage and under that stress type. This standard reflects the reasonable range of phenotypic performance under this stress. If there is no stress, the standard under normal conditions is invoked. This approach achieves dynamic adaptation and refinement of the verification benchmark. It acknowledges that the reasonable performance of the same growth stage varies under different environments, thus providing a judgment standard that best fits the actual situation for subsequent rationality analysis. In this embodiment, an environmental context awareness mechanism is introduced, thereby enabling dynamic and differentiated selection of judgment criteria. This addresses the problem in the background technology where existing rigid data verification models, by ignoring environmental interactions, may misjudge valuable stress physiological data as invalid data. In this scheme, crop phenotypes are the result of genotype-environment interaction. This idea is engineered and implemented through a process of "sensing the environment - identifying stress - switching criteria." This enables the system to understand that "shorter plant height" under drought stress may not be an error that needs to be corrected, but rather a key agricultural response that should be faithfully recorded.

[0070] In one embodiment, the step of determining the reasonableness level of the actual value based on a preset deviation threshold includes:

[0071] Based on the identification information of the target crop or target livestock, obtain a predefined stress resilience coefficient that is associated with the current stress environment type;

[0072] Based on the current environmental state information, the stress intensity of the stress environment type is determined;

[0073] Based on the stress elasticity coefficient and the stress intensity, the preset deviation threshold is dynamically adjusted to obtain an adjusted deviation threshold that adapts to the current environment and variety.

[0074] Based on the adjusted deviation threshold, the reasonableness level of the actual value is determined.

[0075] As described above, Step 1, based on the specific identifier of the currently collected object, queries the variety or species characteristic knowledge base for its preset resilience coefficient for the identified stress environment type. This coefficient quantifies the inherent tendency or tolerance of the organism to vary its phenotypic expression when facing specific adversity. This achieves variety-based precision and personalization of the judgment criteria. The system can distinguish the differences in response to adversity among different genetic backgrounds, for example, knowing that one wheat variety is drought-resistant while another is less so. Step 2 analyzes and calculates the acquired raw environmental data, transforming it into a quantitative indicator characterizing the severity of the current stress. For example, the intensity level of drought is classified according to the number of days when soil moisture content is continuously below a threshold. This achieves quantitative assessment and classification of environmental impact. It advances the qualitative judgment of "drought exists" to the quantitative description of "moderate drought" or "severe drought," enabling the system to perceive the severity and urgency of adversity and provide precise input for dynamic adjustment. Step 3 combines the resilience coefficient characterizing intrinsic tolerance with the stress intensity characterizing the magnitude of external pressure, and dynamically calculates a new deviation threshold applicable to the current specific situation through a preset algorithm model. When tolerance is poor and stress is strong, the threshold may be significantly relaxed. This achieves intelligent adaptation and flexibility in the judgment criteria. It allows the system's fault tolerance to change in real time with variety characteristics and environmental pressure, thus logically acknowledging that "it is normal for vulnerable varieties to grow poorly under severe adversity," avoiding the mechanical application of fixed standards. Step 4 ultimately uses a context-adjusted dynamic threshold, rather than a fixed initial threshold, to assess the degree of deviation from the actual observed data and determine its reasonableness level accordingly. This completes the ultimate reasonableness judgment based on contextual awareness. This ensures that the final judgment considers both the observed value itself and deeply integrates the variety background and environmental context that produced the value, thus arriving at a conclusion that best fits the actual agricultural situation. In this embodiment, the core agronomical concept of "genotype × environment" interaction is further introduced, and the dynamic and personalized adaptation of the judgment threshold is achieved through the combination and calculation of two quantifiable parameters: "elasticity coefficient" and "stress intensity." This invention addresses the problem of existing rigid verification models in the background technology, which completely ignore the quantitative interaction between variety differences and environmental influences. These models are highly likely to misclassify reasonable anomalies caused by environmental stress, which have significant biological value, as invalid, leading to the loss of crucial agricultural information. A simulated intelligent agricultural decision-making model is constructed. This model recognizes that judging the reasonableness of data cannot be separated from the fundamental premise of "who is growing under what environment." By dynamically adjusting thresholds, it essentially performs a high-level anomaly detection, not simply detecting deviations, but determining whether the deviation is within the expected range after understanding its potential causes. Therefore, this invention represents a key evolution from rule-based verification to model-based intelligent scenario assessment.It ensures that the system can both keenly capture real errors in the complex and ever-changing reality of the field and wisely accept and retain "reasonable anomalies" that record the story of the interaction between organisms and the environment, thereby greatly enhancing the scientific value and practical effectiveness of the collected data.

[0076] In one specific embodiment, a field survey at a maize breeding base is used as the scenario. Breeding technician Zhang is conducting a survey on the tasseling traits of maize material numbered "B73" in an experimental field. Holding the equipment, Zhang verbally states the observation results: "B73, field number 5, plant height approximately 2.4 meters, ear height 90 centimeters." The system obtains real-time information through an IoT interface: the field has experienced 15 consecutive days of no effective rainfall, the soil moisture sensor shows a water content of only 45% of field capacity, and the daily temperature is 35℃. Step A: Determine the scenario and establish a dynamic mapping relationship. The system recognizes keywords such as "maize," "tasseling period," and "B73" in the speech, determining the scenario as "maize tasseling trait survey." Subsequently, based on the "B73" variety and "tasseling period," combined with current environmental information (continuous drought, high temperature), the system establishes a dynamic physiological indicator mapping relationship from the knowledge base. This mapping relationship not only includes conventional tasseling period reference values ​​but also specifically relates to the expected trait range under the "moderate moisture-high temperature combined stress" model. Step B: Intelligent Conversion and Initial Data Generation. The system converts speech to text and maps "plant height" and "ear position" to standard fields. "2.4 meters" is converted to "240 centimeters," and "90 centimeters" is confirmed as "90 centimeters," generating initial structured data: {Variety: "B73", Growth Period: "Heading Stage", Plant Height: 240, Unit: cm, Ear Position: 90, Unit: cm}. Step C: Environmental Perception and Differentiated Standard Extraction. The system analyzes the acquired soil moisture and temperature data to determine if a predefined "moisture-high temperature combined stress" type exists. Therefore, instead of using conventional standards, the system extracts the reference value range for the "B73" variety during the heading stage under this specific stress environment from the mapping relationship. For example: Reference range for plant height during the heading stage under conventional conditions: [220, 260] cm. Reference range for plant height under moderate moisture-high temperature stress: [200, 240] cm (expected growth inhibition). Step D: Quantitative Analysis and Grading Judgment. The system compares the actual values ​​with the differential standard: Plant height: The actual value is 240 cm, which is at the upper limit of the stress environment reference range [200, 240] cm. Deviation calculation: For plant height, the value has reached the upper limit of the stress range, and the deviation is low. Reasonableness level judgment: Based on the calculation, the system determines that the plant height data is "reasonable" after considering environmental stress. Although this 240 cm data is also reasonable under the conventional standard, the system's understanding is different at this point: it recognizes that this value was obtained under adversity and has the potential value of indicating the drought resistance of the variety. The system confirms the validity of the data, generates the final structured record, and automatically adds environmental labels: {Environmental context: "water-high temperature combined stress", data confidence status: "reasonable - inhibited by the environment"}.Comparative Demonstration: If a rigid verification method ignoring the environment is used, the system will only use the conventional standard [220, 260] cm for judgment. Although it will not report an error, it will completely lose the crucial agricultural information that "the plant height was measured under drought conditions," greatly reducing the value of the data. Advanced Scenario (If Data Abnormal): Suppose Mr. Zhang verbally states that the plant height is "1.8 meters" (180 cm). The system calculates and finds that this value is significantly lower than the lower limit of the stress interval, with a large deviation, and judges it as "questionable" or "unreasonable." At this time, the system can reason based on physiological mapping relationships: Under such drought conditions, is the extremely low plant height accompanied by other visible symptoms? Then, it can actively guide Mr. Zhang to confirm via voice: "The system has detected that the current drought is severe, and the plant height of 180 cm that you reported is significantly lower than expected. Please confirm whether you have observed severe leaf curling or yellowing of the lower leaves?" This achieves knowledge-based interactive verification.

[0077] In one feasible embodiment, in the field of intelligent agricultural data acquisition, existing technologies for data verification through environmental perception and preset rules, while adaptable to the influence of single environmental factors to a certain extent, still have significant limitations when facing the more common and complex multidimensional and dynamic scenarios in real agricultural production. Besides the identified deficiency of insufficient modeling of the nonlinear superposition effects of multiple stress factors, existing methods face other key challenges in practical applications. For example, the system's heavy reliance on "typical" knowledge models built from historical data makes it difficult to effectively identify and handle breakthrough phenotypic traits that exceed current cognitive scope during breeding, risking misjudging "innovation" as "error." Simultaneously, in data-sparse newly planted areas or niche crop varieties, the system faces a "cold start" dilemma due to a lack of sufficient prior knowledge, leading to the failure of its intelligent verification function or a significant reduction in reliability. Furthermore, existing frameworks do not adequately consider the time-lag effects and historical influences of environmental stress, potentially misjudging reasonable current phenotypes resulting from past events as abnormal.

[0078] In summary, the method further includes: determining whether a predefined stress environment type exists based on the current environmental state information, including:

[0079] From the current environmental status information, at least two coexisting stress factors and their respective intensity values ​​are extracted;

[0080] Query the predefined multi-factor interaction effect knowledge base to obtain the interaction effect coefficients between at least two stress factors;

[0081] Based on the intensity values ​​of each stress factor and the interaction effect coefficient, a comprehensive stress index is generated through weighted calculation.

[0082] The comprehensive stress index is matched with a preset threshold range to determine the corresponding stress environment type.

[0083] As described above, comprehensive analysis of environmental sensor data (such as soil moisture, air temperature, and nutrient concentration) identifies multiple environmental factors that simultaneously exceed their respective normal thresholds and pose potential stress to crop growth, and quantifies the degree of deviation for each factor. This achieves a refined deconstruction and quantitative perception of complex field environmental conditions. The system no longer merely seeks a single "dominant stress," but can comprehensively identify and quantify the specific intensity of multiple coexisting factors such as "drought," "low temperature," and "nitrogen deficiency," providing precise input data for understanding compound stresses. It accesses a knowledge base storing agricultural research results, which defines possible interactions (such as synergy and antagonism) between different combinations of stress factors in the form of data or rules. Based on the identified stress factor combinations, the system queries and obtains the corresponding interaction effect coefficients. It introduces the ability to recognize the nonlinear interactions between stress factors. This enables the system to understand that the harm of "drought superimposed with high temperature" may be far greater than the sum of their individual effects, and that "moderate drought and low temperature" may produce a certain antagonistic effect, thus logically transcending the rudimentary models of simple parallel or weighted averages. Using an embedded mathematical model, the intensity values ​​of each factor are used as basic inputs. Their weights in the comprehensive calculation are adjusted based on the interaction effect coefficient, or nonlinear terms are introduced, ultimately calculating a single quantitative index characterizing overall stress pressure. This achieves the transformation from multiple independent environmental parameters to an integrated, comparable measure of stress pressure. This index comprehensively reflects the "total pressure" exerted on crops by multiple factors and their interactions, providing a unified and scientific benchmark for subsequent judgments. The calculated comprehensive stress index is compared with preset index threshold ranges in the knowledge base corresponding to different stress types (such as "mild combined stress" and "severe hydrothermal stress"), thereby classifying continuous index values ​​into the most matching discrete stress environment type. This completes the mapping from continuously quantified environmental pressure to discrete, actionable business tags. This ensures that subsequent processes (such as calling corresponding reference standards) can proceed based on the most accurate classification of the current combined stress scenario.

[0084] In one embodiment, taking the sampling and grading of "Sunshine Rose" grapes in a fruit warehouse as an application scenario, this invention utilizes the deep understanding capabilities of a large language model in non-growth observation scenarios to achieve intelligent conversion and verification of complex feature descriptions into structured data. User voice input: A quality inspector, holding a handheld device, verbally describes the inspection results for a batch of grape samples: "The variety is Sunshine Rose, the bunch weight is about 500 grams, but the uniformity of the berries is average, the sugar content is about 18 degrees Brix, there are some downy mildew spots, and the coloring is considered excellent." System action: After receiving the voice data, the system, based on keywords such as "variety," "bundle weight," "uniformity," "sugar content," "spots," and "coloring" in the speech recognition text, uses the scenario classification capabilities of the large language model to determine that the current scenario is "post-harvest quality inspection of fruits and vegetables," and calls the predefined collection field set for this scenario. This set not only includes numerical fields such as "bundle weight (g)" and "sugar content (Brix°)," but also enumeristic or graded fields that need to be determined from descriptive language, such as "berry uniformity," "disease status," and "coloring grade." AI-Driven Intelligent Conversion and Initial Data Generation: Field Mapping and Semantic Error Correction: The system utilizes a large language model to perform deep semantic analysis on the identified text. For example, the colloquial phrase "has some downy mildew spots" is accurately mapped to the fields "Disease Type: Downy Mildew" and "Disease Severity: Mild." Simultaneously, the model can associate the vague description "average" with the standard option "Berry Uniformity: Medium." Numerical and Grade Standardization: Based on field definitions, the system standardizes "approximately 500 grams" to "Hill Weight: 500g," and "approximately 18 degrees" to "Sugar Content: 18.0 Brix°." For uncertain expressions like "Coloring is considered excellent," the model can combine context to determine it as "Coloring Grade: Excellent" and assign a confidence level. AI-Powered Deep Validation and Intelligent Interaction: Cross-Field Logic Validation: The system calls upon the agricultural knowledge rules built into the large language model to validate the initial structured data. For example, the model infers that if "Disease Severity" is "mild," then the "Overall Rating" field should not be "poor." The system detected that the user had not provided an "overall rating," triggering a warning for incomplete logic. Reasonableness analysis and candidate generation: The system analyzes the characteristics of the "Sunshine Rose" variety based on an AI model. Assuming the sugar content of this variety is typically higher than 17 degrees Brix, the reported 18 degrees Brix is ​​reasonable. Simultaneously, the model can generate intelligent follow-up questions for missing or questionable fields, such as: "Based on your description of a sugar content of 18, excellent coloring, and minor disease, the system infers an overall rating of 'Excellent.' Please confirm. Also, please provide information on the freshness of the fruit stems." Interactive correction and confirmation: The quality inspector can respond via voice: "Overall rating confirmed as Excellent, fruit stems are bright green." The system completes data completion and correction through voice interaction, forming the final structured data record.This embodiment demonstrates how, in a general agricultural quality inspection scenario that does not involve growth stages or environmental stresses, this invention, through deep integration of a large language model, accurately transforms vague and non-standard colloquial descriptions such as "average," "somewhat," and "rather good" into standardized field values. Utilizing built-in variety characteristics and quality logic rules, it automatically identifies data contradictions and omissions and proactively initiates intelligent follow-up questions. Data confirmation, correction, and completion are quickly completed through voice dialogue, greatly improving the efficiency and accuracy of inputting complex phenomenological descriptions.

[0085] Reference Figure 3 This application also provides a voice-based intelligent agricultural structured data acquisition system, comprising:

[0086] Receiving module 1 is used to receive voice data related to agriculture input by the user;

[0087] Module 2 is invoked to determine the current data acquisition scenario based on the voice data and to invoke a predefined set of acquisition fields for the acquisition scenario, wherein the set of acquisition fields includes the standard name, data type, and value constraints of the fields;

[0088] Recognition module 3 is used to perform text information recognition on the voice data;

[0089] The generation module 4 is used to match the identified text information with the collection field set according to the recognition result, map the expression in the text to the corresponding standard field name, and standardize the numerical value and unit in the text according to the data type and value constraints to generate initial structured data.

[0090] The verification module 5 is used to verify the initial structured data based on the value constraints in the collection of fields and the preset agricultural knowledge rules.

[0091] Module 6 is used to determine the final structured data based on the verification results.

[0092] As described above, it is understood that each component of the voice-intelligent agricultural structured data acquisition system proposed in this application can realize the function of any of the voice-intelligent agricultural structured data acquisition methods described above, and the specific structure will not be repeated.

[0093] Reference Figure 4 This application also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 4As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores monitoring data and other data. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a voice-based intelligent agricultural structured data acquisition method.

[0094] The processor described above executes the voice-based intelligent agricultural structured data acquisition method, comprising: receiving agricultural-related voice data input by a user; determining the current data acquisition scenario based on the voice data, and calling a predefined collection field set for the acquisition scenario, wherein the collection field set includes standard names, data types, and value constraints of the fields; performing text information recognition on the voice data; matching the recognized text information with the collection field set based on the recognition result, mapping the expressions in the text to corresponding standard field names, and standardizing and converting the numerical values ​​and units in the text according to the data types and value constraints to generate initial structured data; verifying the initial structured data based on the value constraints in the collection field set and preset agricultural knowledge rules; and determining the final structured data based on the verification result.

[0095] One embodiment of this application also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements a voice-based intelligent agricultural structured data acquisition method, comprising the following steps: receiving agricultural-related voice data input by a user; determining the current data acquisition scenario based on the voice data, and calling a predefined collection field set for the acquisition scenario, wherein the collection field set includes standard names, data types, and value constraints of the fields; performing text information recognition on the voice data; matching the recognized text information with the collection field set based on the recognition result, mapping the expressions in the text to corresponding standard field names, and standardizing and converting the numerical values ​​and units in the text according to the data types and value constraints to generate initial structured data; verifying the initial structured data based on the value constraints in the collection field set and preset agricultural knowledge rules; and determining the final structured data based on the verification result.

[0096] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media provided in this application and in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0097] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0098] The above description is only a preferred embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural changes made based on the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for acquiring structured agricultural data based on voice intelligence, characterized in that, The method includes: Receives voice data related to agriculture input by the user; Based on the voice data, the current data acquisition scenario is determined, and a predefined set of acquisition fields for the acquisition scenario is invoked, wherein the set of acquisition fields includes the standard name, data type, and value constraints of the fields; Perform text information recognition on the voice data; Based on the recognition results, the recognized text information is matched with the collection field set, the expression in the text is mapped to the corresponding standard field name, and the numerical value and unit in the text are standardized and converted according to the data type and value constraints to generate initial structured data. Based on the value constraints in the collection field set and the preset agricultural knowledge rules, the initial structured data is validated. Based on the verification results, the final structured data is determined; The method also includes establishing a physiological indicator mapping relationship matching the growth stage or growth cycle of the target crop or livestock corresponding to the collection scenario; performing a rationality analysis on the fields involving physiological or growth status in the initial structured data based on the physiological indicator mapping relationship; and generating candidate correction data based on the physiological indicator mapping relationship if the analysis result is unreasonable, and guiding the user to confirm or correct it through voice interaction. Based on the physiological indicator mapping relationship, a rationality analysis is performed on the fields related to physiological or growth states in the initial structured data. The steps include: extracting reference value ranges or reference value features for each physiological state field corresponding to the reproductive stage or growth cycle from the physiological indicator mapping relationship; comparing the actual values ​​of the same physiological state field in the initial structured data with the corresponding reference value ranges or reference value features; calculating the deviation between the actual values ​​and the corresponding reference standards based on the comparison results; and determining the rationality level of the actual values ​​based on a preset deviation threshold, wherein the rationality level includes at least reasonable, questionable, and unreasonable. The step of extracting reference value ranges or reference value features of each physiological state field corresponding to the reproductive stage or growth cycle from the physiological indicator mapping relationship includes: obtaining current environmental state information associated with the target crop or target livestock; determining whether a predefined stress environment type exists based on the current environmental state information, wherein the stress environment type is used to characterize a set of environmental conditions that have a specific pattern of influence on the growth and development of the target crop or target livestock; if it exists, extracting reference value ranges or reference value features that are jointly corresponding to the reproductive stage or growth cycle and the stress environment type from the physiological indicator mapping relationship; if it does not exist, extracting reference value ranges or reference value features that correspond to the reproductive stage or growth cycle and are under normal conditions from the physiological indicator mapping relationship.

2. The method for acquiring structured agricultural data based on voice intelligence according to claim 1, characterized in that, The step of validating the initial structured data based on the value constraints in the collection field set and preset agricultural knowledge rules includes: Obtain the preset agricultural knowledge rules, wherein the agricultural knowledge rules define the logical constraint relationships between different data collection fields; The values ​​of the relevant fields in the initial structured data are substituted into the corresponding logical constraint relationships for calculation; Based on the calculation results, determine whether the initial structured data satisfies the logical constraint relationship; In response to the judgment being satisfied, the initial structured data is determined to have passed the verification; In response to the determination that the condition is not met, the initial structured data is marked or a correction is triggered.

3. The method for acquiring structured agricultural data based on voice intelligence according to claim 1, characterized in that, The steps include: matching the identified text information with the collected field set, mapping the expressions in the text to corresponding standard field names, and standardizing the numerical values ​​and units in the text according to the data type and value constraints to generate initial structured data. Based on the standard names of each field in the collection of fields and the predefined word list, the words in the text information are matched with the corresponding fields; Extract the numerical and unit descriptions corresponding to each field from the matched text information; Based on preset conversion rules, the numerical description is converted into a standard numerical format that conforms to the data type; Based on the units defined for each field in the collection field set, the unit descriptions are converted or transformed into the defined units; Based on the standard numerical format and the converted or transformed units, the values ​​of each field in the initial structured data are generated.

4. The method for acquiring structured agricultural data based on voice intelligence according to claim 1, characterized in that, The step of determining the reasonableness level of the actual value based on a preset deviation threshold includes: Based on the identification information of the target crop or target livestock, obtain a predefined stress resilience coefficient that is associated with the current stress environment type; Based on the current environmental state information, the stress intensity of the stress environment type is determined; Based on the stress elasticity coefficient and the stress intensity, the preset deviation threshold is dynamically adjusted to obtain an adjusted deviation threshold that adapts to the current environment and variety. Based on the adjusted deviation threshold, the reasonableness level of the actual value is determined.

5. A voice-based intelligent agricultural structured data acquisition system, used in the method described in any one of claims 1-4, characterized in that, include: The receiving module is used to receive voice data related to agriculture input by the user; The calling module is used to determine the current data acquisition scenario based on the voice data, and to call a set of acquisition fields predefined in the acquisition scenario, wherein the set of acquisition fields includes the standard name, data type and value constraints of the fields; The recognition module is used to perform text information recognition on the voice data; The generation module is used to match the identified text information with the collection field set based on the recognition results, map the expressions in the text to the corresponding standard field names, and standardize and convert the numerical values ​​and units in the text according to the data type and value constraints to generate initial structured data. The verification module is used to verify the initial structured data based on the value constraints in the collection of fields and the preset agricultural knowledge rules. The determination module is used to determine the final structured data based on the verification results; It also includes establishing a physiological indicator mapping relationship matching the growth stage or growth cycle of the target crop or livestock corresponding to the collection scenario; performing a rationality analysis on the fields involving physiological or growth status in the initial structured data based on the physiological indicator mapping relationship; and generating candidate correction data based on the physiological indicator mapping relationship if the analysis result is unreasonable, and guiding the user to confirm or correct it through voice interaction. Based on the physiological indicator mapping relationship, a rationality analysis is performed on the fields related to physiological or growth states in the initial structured data. The steps include: extracting reference value ranges or reference value features for each physiological state field corresponding to the reproductive stage or growth cycle from the physiological indicator mapping relationship; comparing the actual values ​​of the same physiological state field in the initial structured data with the corresponding reference value ranges or reference value features; calculating the deviation between the actual values ​​and the corresponding reference standards based on the comparison results; and determining the rationality level of the actual values ​​based on a preset deviation threshold, wherein the rationality level includes at least reasonable, questionable, and unreasonable. The step of extracting reference value ranges or reference value features of each physiological state field corresponding to the reproductive stage or growth cycle from the physiological indicator mapping relationship includes: obtaining current environmental state information associated with the target crop or target livestock; determining whether a predefined stress environment type exists based on the current environmental state information, wherein the stress environment type is used to characterize a set of environmental conditions that have a specific pattern of influence on the growth and development of the target crop or target livestock; if it exists, extracting reference value ranges or reference value features that are jointly corresponding to the reproductive stage or growth cycle and the stress environment type from the physiological indicator mapping relationship; if it does not exist, extracting reference value ranges or reference value features that correspond to the reproductive stage or growth cycle and are under normal conditions from the physiological indicator mapping relationship.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images